Olivier Fercoq

Home Biography Publications Talks CV Software


My CV

Employment

  • 2015-present: Assistant professor (maître de conférences), Telecom ParisTech, France, Signal and image processing department.
  • 2014-2015: Postdoctoral research assistant, Telecom ParisTech, France (P.I. Pascal Bianchi), Signal and image processing department.
    Focus: Randomized optimisation algorithms for machine learning
  • 2012-2014: Postdoctoral research assistant, University of Edinburgh, UK (P.I. Peter Richtarik), Edinburgh Research Group in Optimization
    Focus: Parallel coordinate descent methods for general convex functions

Education

  • 2009-2012: Ph. D. in Applied Mathematics, Ecole Polytechnique (Paris, France) Defended on 17th September 2012 (with distinction)
    Supervisors: Stephane Gaubert, Marianne Akian and Mustapha Bouhtou (Orange Labs)
    Thesis: Optimization of Perron eigenvectors and applications: from web ranking to chronotherapeutics
    Funded by a research contract between Inria and Orange Labs
  • 2008-2009: Master of research, Paris 6 University (France): Optimization, Game Theory and Modelling in Economy, (with distinction)
    6 months internship, EDF R & D, Clamart (France): Elaboration of a strategy of valuation for intra-day ajustement offers in the electricity market
  • 2006-2009: Engineer's degree, ENSTA-ParisTech (France), major: Simulation and mathematical engineering, specialisation: Optimization and Control
    3 months internship, Inria Paris-Rocquencourt: Synthesis of a control law for a vehicle following system

Prizes

  • 2015: Second prize of the IMA Leslie Fox Prize in numerical analysis.
  • 2013: Ph. D. Thesis Award of the Gaspard Monge Programme for Optimisation and Operational Research (best Ph. D. Thesis in optimisation and operational research in France)

Teaching experience

  • Since 2015: courses, tutorials and projects in optimization, Fourier analysis, machine learning and signal processing
  • 2015: Supervision of one Master's degree internship: Eugene N'Diaye (Dynamic variable screening for supervised learning in large dimensions with structured sparsity constraints)
  • 2013: Supervision of two MSc summer projects: Andrew Cairns (Web advertisement pricing) and Yun Yang (Maximizing the Revenue Considering Customer Choice)
  • 2013: The University of Edinburgh. Optimization Methods in Finance: managed the lecture and computer lab on robust optimization and conic optimization (50 MSc students) Computing and Numerics: teaching assistant for computer labs (15 2nd-year students) Introduction to linear algebra: teaching assistant (12 1st-year students)
  • 2012: ENSTA-ParisTech, Teaching assistant: Quadratic optimization (15 3rd-year students) and Initiation to finite elements method (15 4th-year students)

Software

  • 2014: Contribution to Scikit-learn (Python): Accelerated coordinate descent for the Lasso
  • 2013: OpenMP/C++ code: Accelerated Parallel PROXimal coordinate descent (APPROX). Prototype solved SVM, Lasso, $l_1$ regression with millions examples and features.
  • 2011: Scilab code for the optimization of the Perron value (chemotherapy) or the Perron vector (web page ranking)
  • 2010: Deliverable of the research contract with Orange Labs : Matlab code with a mexfile interface with the WebGraph library (webgraph.dsi.unimi.it) for the optimization of PageRank. The prototype enabled us to deal with graphs up to 100 million nodes.

Languages

  • French: Mother tongue
  • English: Fluent
  • German: Fluent
  • Chinese: Basic knowledge in writing and speaking
Design by Marion Chagne-Fercoq